Eurorack DIY: Quantizer Update (Episode 2)
Вставка
- Опубліковано 29 сер 2024
- Here's an update progress of the Quantizer project. We will continue the design, have a look into the RPi Pico ADC and my thoughts around the User Interface. We will also explore the calibration features.
Useful Links:
www.allaboutci...
wiki.openmusicl...
pico-adc.marko...
You can find all of my designs on my GitHub:
github.com/tki...
Some modules that I have designed myself are available in my Tindie shop:
www.tindie.com...
Thank you for watching!
#eurorack #synthdiy
This is looking great Tommy! 👊
Thank you! Cheers!
Great video as usual Tommy! The Pi Pico and the adc... Disable the buck converter pin 37 3.3V_EN and provide a clean 3.3V. Then the adc work as it should!
Yes, I will make that change I think. I have seen it mentioned on the RPi forum.
@@MeeBilt I have tested it and I get 10 stable bits and 2 noisy bits. They get stable as well if you average 8 or 16 samples.
Awesome video
Thanks! Glad you enjoyed it!
Hi! Thanks for these videos, they're helping me with my own design (I'm using a DAC instead of the PWM method). I'm very curious to see your code for scales etc, because I'm having a pretty hard time figuring out how to do that efficiently.
Hello, I will push my code to Github soon, but in the mean time there are a couple of projects that are already open source that you can look at:
github.com/alicedb2/tinyQuan/tree/main
github.com/PierreIsCoding/sdiy/tree/main/Quantizer
I have been using a similar approach as the second example for my design.
the Pico WANTS perfect square signals for triggers, now anything over 0.65 volts triggers the pico, you should set the threshold like 2.5 Volts.
You really need to use: LM393, LM339 or Schmidt trigger transistor setup.
Schmidt trigger transistor on wiki and use Falstad simulator to figure out the perfect resistor setup using the LEAST mount of current while maintaining the functionality but doesn't have threshold. that's why LM393 (LM339) is best, also allows voltage conversion from 10V+ to 3v33, because the comparator is open drain.
Yes, I will look into that, thanks for pointing it out. I will make some updates to the design anyway so I will fix the trig inputs at the same time. LM393 seems like a good choice here since I have two trigger inputs.
@@MeeBilt Yeah, Use LM4040 with 100nF to create 2.5V threshold. your current setup allows any weird voltage that shows up to be accepted, albeit it works its not what you want, potentially making or causing havoc, not good for you if you sell these.
@@MeeBilt Mutable Instruments, when they where active, they build a specialized testing PCB with software for that specified module in question and have it create any possible combination of voltages & what not and have that run for like 7 days non stop, to be sure it does what it's supposed to do, until the module passes, it could take up to 2 years before it came to market.
Hur många kort beställde du ?
Jag är klart intreserad av ett och kan stötta med lite kod
Hej, jag kommer göra ett nytt kortvarv för att fixa några problem och PCB filer och källkod kommer att finnas på mn Github inom kort.
Skicka ett mail via YTs kontakt-funktion om du är intresserad av ett av de gamla korten, så kan vi nog lösa det :)
1:04 much better if LM393, LM339 or Schmidt trigger transistor setup for triggers.
Also much better to use 16 bit Texas Instruments ADC rather these onboards, like: ADS8339
Sample Rate: 250 kHz, datasheet shows how to build dedicated power supply for analog input, input opamp config then use ADG419 to switch between inputs to sample.
I will probably stick to the built in ADC for this module, but for the next one I will probably need more analog inputs so I will go for external ADC.
@@MeeBilt try 'ADS8339' see the datasheet for analog power supply and reference voltage designs for this chip, With a low ohm switch IC, like: ADG5208F, ADG5404F, ADGS1408. Analog Devices has analog switch IC's with like 2.5 Ohm resistance for such applications if you want precision, even with SPI. just a day searching and looking datasheets.
: )
You could cut tracks on your current PCB and insert the new options.
@@MeeBilt Just hack you current board, desolder some resistors to use the pads for soldering cables, design tiny PCB with comparators for trigger inputs, design another with ADC, dedicated power supplies and switches / multiplexer, add emough test points, some additional spots for optional capacitors.
You'll be cheaper, than find out you have another PCB that aint right, i did this as well once on module i was figuring out.
In your case it's software, you need to know the principle of "state machine", very important. it's like PLC logic used in factories. most software out there, provides the functional, but not really the brains.
This kind of logic you design on paper, there is or are an ISO definition for that, once you done on paper, you make the software and from the get go it works and is bug free, just in one compile, the first one. there are symbols for that.
@@AnalogDude_ Yep, that is basically how I go about testing design-changes. I normally nail it in 2 board-spins or so.
I've been using state machine based designs for many years in embedded systems. It's great for implementing UI and complex control functions for example.
@@MeeBilt Ya, you don't need update a display millions of times a second or check buttons that often. most do not know, the ADC ic i suggested can have 2.5V reference Voltage and you can divide theoretical 10V input signal by 4, easier input opamp configuring, like 10k / 2.49K + 1 Ohm. using 0.1% resistors or better.